Document key: consent_ai_ocr_processing
Artificial Intelligence and OCR Processing Notice
The data scope and limits of user-facing, professionally assisted, automated, or safety-related artificial-intelligence flows and document-reading operations if enabled.
- Version
- 2.0
- Last updated
- July 25, 2026
- Data controller
- EGC YAZILIM SAĞLIK HİZMETLERİ LTD. ŞTİ.
Purposes and processing types
Artificial intelligence may be used in safety and symptom-assessment tools where a user submits a question, text, document, or image; daily tips and weekly evaluations; doctor-response drafts and message translation; forum-assistant, content-draft, or forum-seed preparation; automated moderation of community posts and direct messages; and background or queued operations. Accordingly, not every operation begins with a separate screen action, and some flows may run automatically to provide security, moderation, or a feature enabled by the user. The applicable health-data condition and feature terms must apply, and a user action does not replace a separate affirmative choice where one is required before personal or health data is transferred to a third-party provider.
Information processed and safeguards
- Depending on the flow, a question, message, community content, selected file or image, translation text, pregnancy context, gestational week, measurements, symptoms, lifestyle, medication or supplement, triage and trend data, and a technical user, content, job, or operation identifier may be processed. A doctor or admin draft may use the limited clinical or operational context required by the relevant request.
- Data sent to a provider may include a name, phone number, email address, Turkish Republic identification number, or another direct identifier that the user included in a question or document, as well as a technical user or job identifier used in a weekly flow. Users should remove identity and contact fields that are unnecessary for the purpose; data minimization and access limits are applied according to the nature of the relevant flow.
- Output may be inaccurate, incomplete, or outdated. An automated result is not medical diagnosis, prescription, treatment, or emergency assessment.
- Artificial-intelligence or OCR providers may be used only to deliver the relevant function, protect security, and monitor service quality to the extent necessary. Provider categories, processing locations, retention approaches, possible human access, and model-training conditions are assessed together with current vendor records and feature notices. Health data may not be used for targeted advertising, profile sales, or a provider's independent marketing.
Human review, transfers, and choice
A doctor-response draft or document summary may be subject to human review in the appropriate workflow before it is presented as a definitive medical response. Users should rely on the original document and professional assessment and must not wait for artificial-intelligence output in an emergency.
Where provider infrastructure causes international processing, an adequacy decision or appropriate safeguard must be established and verified before processing begins. Explicit consent may be considered separately only for an occasional transfer that meets the statutory conditions. A user may submit a request concerning future optional processing by not using an optional flow, disabling an available feature control, or contacting support. Processing based on another legal ground, such as necessary security or community moderation, and limited legal, security, and dispute records may continue for the required scope and period.